Acupuncture in Migraine Prevention- A Randomized Sham Controlled Study With 6-Months Posttreatment Follow-up

Acupuncture in Migraine PreventionA Randomized Sham Controlled Study With 6-months

Posttreatment Follow-up

Jerusa Alecrim-Andrade, MD, MSc,*w Jayme Antunes Maciel-Junior, MD, PhD,w zXavier Carne`, MD, PhD,y Geraldo Magela Severino Vasconcelos, BSc,w

and Heleno Rodrigues Correa-Filho, MD,w J

Objective: To assess the ecacy of acupuncture in migraine

prophylaxis.

Methods: Thirty-seven patients with migraine were enrolled in a

randomized control trial at the Headache clinic located in a

University Hospital. Real and sham acupuncture groups

received 16 acupuncture sessions over 3 months. Treatment

was individualized in the real acupuncture group and minimal

acupuncture was used in the sham group. The primary end point

was the percentage of patients with a Z50% reduction in theirmigraine attack frequency in the second, third, fourth, fth, and

sixth (months) compared with the rst one (baseline period).

Primary and secondary end points were measured comparing

headache diaries.

Results: Real acupuncture group showed improvement with

signicant dierences compared with the sham acupuncture

group in the primary ecacy end point (P=0.021) at the

second month of the treatment. Dierences also appeared in 2

secondary end points: number of days with migraine per month

(P=0.007) in the second month and the percentage of patients

with Z40% reduction in migraine attack frequency in the rst(P=0.044) and second months (P=0.004) of the treatment.

These dierences disappeared in the third (last) month of the

treatment as a consequence of the high improvement of the

sham acupuncture group. Comparisons within each group

showed that several migraine parameters evaluated improved

signicantly in both groups.

Conclusions: Individualized treatment based on traditional

Chinese medicine plays a role in preventing migraine attacks.

Nevertheless, sham acupuncture had similar eects. Major

conclusions were limited by the small sample sizes however the

observed trends may contribute to design future trials.

Key Words: acupuncture, migraine, prophylaxis, randomized

controlled trial, headache, ecacy, eectiveness

(Clin J Pain 2008;24:98105)

The history of acupuncture shows the very old relation-ship between this medical approach and the treatmentof pain.1 In the western world, acupuncture is widely usedand accepted for controlling chronic pain.2,3 Up to now,according to the Evidenced Based Medicine, acupuncturehas only showed uncontested eectiveness in pain in 2areas: chronic low back pain and postoperative dentalpain.46

The prevalence of migraine in the United States isabout 18% for females and 5% for males.7 These rates liketo be lower in Asian Americans population as well as inAsian and African countries.810 Migraine is not recog-nized as a specic disease in the theory of traditionalChinese medicine (TCM).11 Nevertheless, acupuncture hasbeen largely used for people with migraine in westerncountries.12 In the last 30 years, several trials wereperformed to test acupuncture ecacy to controlmigraine.2,1322 However, a series of methodologicalproblems make it impossible to reach denitive conclu-sions.2325 Recently published trials have noted thatacupuncture plays a role in migraine2628 and tension-typeheadache,29 but they showed no statistically signicantdierences between acupuncture based on principles ofTCM and the sham acupuncture. The aims of the presentstudy were to assess the direction and magnitude ofacupuncture eects in migraine prophylaxis.

STUDY DESIGN

Patients and MethodsThis study was conducted between December 2001

and June 2003 in the State University of Campinas(UNICAMP), Brazil. The Institutional Ethics Committeeapproved it in June 2000 in accordance with theCopyright r 2008 by Lippincott Williams & Wilkins

Received for publication April 8, 2007; revised June 15, 2007; acceptedAugust 17, 2007.

From the *Department of Medicine, Autonomous University ofBarcelona; yClinical Pharmacology Unit, Hospital Clinic, Barcelona,Catalunya, Spain; zDepartment of Neurology, Headache Clinic,School of Medical Sciences; JEpidemiology Faculty, Department ofSocial and Preventive Medicine, School of Medical Sciences; andwState University of Campinas, Campinas, Brazil.

Supported by a grant from the State of Sao Paulo Research Foundation(FAPESP, grant no. 00/09985-0). ISRCTN 93327878. Competinginterests: None.

Reprints: Jerusa Alecrim-Andrade, MD, MSc, Rua Otavio Machado,225 apto/33 Campinas-SP 13.076-160, Brazil (e-mail: [email protected]).

ORIGINAL ARTICLE

98 Clin J Pain Volume 24, Number 2, February 2008

Declaration of Helsinki. Patients were recruited via anadvertising campaign in the media. The protocol wasstructured according to the Guidelines for ControlledTrials of Drugs in Migraine30 and the Guidelines forClinical Research on Acupuncture.31

Thirty-seven consecutive patients with migrainewith or without aura for at least 1 year diagnosedaccording to the International Headache Society32 criteriawere included. Major inclusion criteria were age between18 and 50; 2 to 6 migraine attacks per month (baselineperiod) and having not used prophylactic migraine drugsor acupuncture in the last 3 months. The principalexclusion criteria were diagnoses of any other painsyndrome including other associated headache type;having used any migraine prophylactic drugs in the 3months before inclusion; having used analgesics for morethan 10 days a month.

Patients were analyzed retrospectively on the basisof their clinical history and prospectively through head-ache diaries and visits by an experienced neurologist.Clinical and demographic data are shown in Table 1. Atthe rst neurologist visit selected patients were informed(orally and through a print-out) about the study, risks,and their right to withdraw at any time without specifyingreasons. A group meeting allowed learning how to keep aheadache diary. The diary required considerable amountof details about the daily behavior of migraine through-out the month, including the intensity in the morning,afternoon, and night, duration, medication intake (typeand doses), nausea, vomiting, and menstruation. Patientsspent a month (baseline period) keeping their diary andwere scheduled for a second neurologists visit. Data frombaseline period allowed for checking inclusion andexclusion criteria. Selected patients provided writteninformed consent. Before the acupuncture treatment,patients were evaluated by an acupuncture specialistphysician who recorded the clinical history and physicalevaluation on the basis of the principles of TCM. It wasfollowed by an active acupuncture treatment period.

RandomizationPatients were randomly assigned to sham or real

acupuncture groups through opaque numbered andsealed envelopes. The random digits list33 was used todetermine dierent letters sequence in 7 blocks. Eachblock contained 3 letters C and 3 letters D. Each letter

signed one of the 2 acupuncture groups. Only the medicalacupuncturist knew the meaning (group) of each letter.

BlindingBoth groups were treated with 16 acupuncture

sessions. The neurologist and the medical acupuncturistwere prohibited from commenting on any treatmentdetails with patients. The medical acupuncturist adopteda uniform, neutral attitude toward the patients so as notto disrupt the blind design of the trial.

At the end of the study, patients were invited tocomplete a questionnaire that evaluated the treatmentthey received and to state their impression about whichtreatment group they thought they had been included.

Outcome MeasurementPatients were instructed to complete diaries for 6

months: baseline period (1mo), acupuncture treatmentperiod (3mo), early follow-up (rst month after the lastacupuncture session), and late follow-up (sixth monthafter the last acupuncture session). Posttreatment follow-up periods were designed to test the long-lasting eects.

The Acupuncture TreatmentThere were several common approaches in both

groups including: bilateral points; usage of disposable,sterile steel needles (0.25mm 40mm); skin disinfectionwith 70% alcohol; needles in place for 30 minutes and nomoxa or electrical stimulation. Treatment could bealtered from session to session, as it happens whenacupuncturists treat patients in their oces.

Real acupuncture treatment was individualizedbased on the principles of TCM. Each treatment wasdened for at least 2 experienced physicians specialized inacupuncture for more than 11 years. A maximum of 20needles were inserted. However, the central principle ofthe treatment was followed. The needles were manipu-lated by rotation methods to produce a characteristicsensation known as De Qi, which was explained to thepatients of this group in the rst session.

Sham acupuncture treatment was minimal acupunc-ture and consisted of a very supercial insertion inacupuncture points with needles almost falling out. Thenumber of needles varied from 10 to 15. No manipulationwas done. The points were selected after an extensiveconsultation of the Chinese acupuncture literature withno references to eects on headaches.3439 Some shampoints were localized in the head to preserve the patientsblinding.

Parameters and StatisticsThe primary end point was the percentage of

patients with Z50% reduction in migraine attackfrequency each month compared with the baseline period.A new attack was recorded when the patient had been freeof headaches for 48 hours before the pain returning.Secondary ecacy parameters included the percentage ofpatients with Z40% reduction in migraine attackfrequency, the number of attacks per month, the numberof days with migraine per month, the total duration of

TABLE 1. Clinical Details of the Population Studied

Real Acupuncture(n=19)

Sham Acupuncture(n=17) P

Mean age (y) 36.79.2 33.29.2 nsSex (F/M) 17/2 15/2 nsSchooling (y) 13.53.5 13.63.5 nsDuration of disease (y) 20.610.8 14.57.6 nsNo. attacks (baselineperiod)

4.50.9 4.71.0 ns

Nausea 18 (94.7%) 17 (100%) nsMigraine without aura 17 (89.5%) 9 (52.9%) nsMigraine with aura 2 (10.5%) 8 (47.1%) 0.025

Clin J Pain Volume 24, Number 2, February 2008 Acupuncture in Migraine Prevention

r 2008 Lippincott Williams & Wilkins 99

migraine pain in hours per month, the mean totalduration of each migraine attack, the mean headacheseverity in each attack, the amount and the type of rescuemedication used per month, nausea frequency (number ofdays with nausea per total of migraine days per month),and vomiting frequency (total of vomiting episodes pertotal of migraine days per month). All of these parameterswere compared within and between groups, with thereference point being the baseline period in each group.The severity of headache was evaluated on a 4-pointscale (0no headache, 1mild headache=migraine notinterfering with daily activities, 2moderate head-ache=migraine interfering with daily activities, and

3severe headache=migraine making normal dailyactivities impossible) following International HeadacheSociety recommendations.30

Statistical comparisons were made using univariateanalysis with the w2 test and Fisher exact test. Compar-isons within each group were made using the nonpara-metric Mann-Whitney test and dierences were estimatedusing the Wilcoxon Signed Rank test. The signicant levelused was P

groups. Thirty-six patients were included in the statisticalanalysis (Fig. 1). For unknown reasons, a patient of thesham acupuncture group did not complete her diary. Asshown in Table 1, both groups were homogeneous inalmost all of the characteristics. An exception was in thetype of migraine where migraine patients with aura weresignicantly more numerous in the sham group.

A patient from the real acupuncture group wasexcluded in the early posttreatment follow-up period,because she started to take amitriptyline for depression.There was an additional dropout in the late posttreatmentfollow-up from the real acupuncture group. The patientmoved to another city and did not send in her last diary.

EfficacyIn the rst and second months of acupuncture

therapy, the frequency of migraine attacks fell by Z50%in greater proportion in the real acupuncture group, andstatistical signicant dierences appeared in the secondmonth of the treatment (P=0.021). In the third and lasttreatment month, the frequency of migraine attacksdeclined in both groups, especially in the sham group.As a result, the signicant dierence seen between groupsin the previous months was no longer detected(P=0.332). During the posttreatment follow-up periods,when the end points were analyzed, no statisticallysignicant dierences were observed between groups(Fig. 2).

Dierences with statistical signicance betweengroups appeared in the second month of the treatmentin 3 secondary parameters: number of days with migraineper month (P=0.006), the total duration of migrainepain in hours per month (P=0.025), and reduction of

Z40% in the frequency of migraine attacks (P=0.004).Greater improvement was observed in the real acupunc-ture group. From the third month of the treatment to thelate posttreatment follow-up, the statistical dierencebetween groups disappeared (Figs. 3, 4).

Comparisons within each group showed that allmigraine pain parameters evaluated improved signi-cantly in both groups except for headache severity.However, that improvement appeared with statisticalsignicance in the real acupuncture group from the rstmonth of the treatment through the late posttreatmentfollow-up. In the sham acupuncture group, the improve-ment started only in the second month of the treatmentand then stayed right through. The evolution of theimprovement within each group for the total duration ofpain in hours is displayed in Figure 5.

Rescue MedicationThere was a reduction in the total intake of rescue

medications in both groups. However, there were nostatistical signicant dierences between them. The typeof rescue medication varied. Most of the patients usedmore than 2 types of medication that included simple andcombined analgesics, nonsteroidal anti-inammatorydrugs, ergots, and triptans.

Associated SymptomsThere were no statistical signicant dierences

between groups in the associated symptoms (nausea andvomiting). In the analysis within each group appeared asignicant and progressive reduction in the frequency ofnausea during migraine attacks in the real acupuncturegroup from the rst month of acupuncture through to the

FIGURE 2. Percentage of responders(reduction 50% in migraine attack fre-quency) in each diary (1mo) comparedwith the baseline period (diary 1) in thereal and the sham acupuncture groups.Diary 2 = first month of the treatment,diary 3 = second month of the treat-ment, diary 4= third month of thetreatment, diary 5 = first month afterthe treatment, and diary 6 = sixth monthafter the treatment. In this figure, the Pvalue is referred to the differencesbetween groups.



late posttreatment follow-up. The same was not observedin the sham group.

Adverse EventThe patients were instructed to report all adverse

events (AEs) to the medical acupuncturist. No serious

AEs were observed in the 576 sessions carried out andabout 8640 needles inserted. The AEs appeared mainly atthe sites where needles were inserted. The most commonevent was ecchymosis. It was reported in only one out of8 sessions without signicant dierence between groups.All AEs were classied as low risk.40

Patients PerceptionsAt the end of acupuncture treatment, patients stated

that they would like to be treated with acupuncture againclassifying treatment as good, very good, or excellent.When asked about what type of treatment they received,the reply was I dont know for 52.5% of patients in thereal acupuncture group and for 64.7% of patients in thesham group. Patients who correctly identied theirtreatment were 36.8% in the real acupuncture and11.76% in the sham group. There was no statisticalsignicant dierence between the replies from the 2groups, indicating that the blinding procedure waseective.

DISCUSSIONMore women than men were studied in our sample.

This may be a consequence of the higher number ofwomen in the population; migraine is more common infemales than in males41 and it is recognized that womenseek medical care for headaches more often than men.42

There was a signicant statistical dierence con-cerning migraine with aura in the sham acupuncturegroup compared with the real acupuncture group.

FIGURE 3. Percentage of patients with reduction of 40% ormore in migraine attacks frequency in the real and shamacupuncture groups compared with the baseline period. Diary2 = first month of the treatment, diary 3 = second month of thetreatment, diary 4 = third month of the treatment, diary5 = first month after the treatment, and diary 6 = sixth monthafter the treatment. In this figure, the P value is referred to thedifferences between groups.

FIGURE 4. Total of migraine days in each diary (1mo) in thereal and the sham acupuncture groups. Diary 1 =baselineperiod, diary 2 = first month of the treatment, diary 3 = secondmonth of the treatment, diary 4 = third month of thetreatment, diary 5 = first month after the treatment, and diary6 = sixth month after the treatment. In this figure, the P value isreferred to the differences between groups.

FIGURE 5. Total duration of pain in hours in each diary (1mo)in the real and the sham acupuncture groups compared withtheir own baseline period. Diary 1=baseline period, diary2 = first month of the treatment, diary 3 = second month of thetreatment, diary 4 = third month of the treatment, diary5 = first month after the treatment, and diary 6 = sixth monthafter the treatment. In this figure, the P value is referred to thedifferences within each group compared with the baselineperiod.

Alecrim-Andrade et al Clin J Pain Volume 24, Number 2, February 2008

102 r 2008 Lippincott Williams & Wilkins

However, up to now, the literature has demonstrated thatthe presence or absence of aura has not aected theoutcome of any applied treatment.43

There were several positive features of this studyincluding: the homogeneity of the sample in almost alldemographic characteristics; the randomized design ofthe trial; the successful blinding of the patients, evaluator,statistician, and research assistants, the very low rate ofdrop-out with less than 10% of patients from thebeginning up to the late follow-up period and nally,there was only one experienced medical doctor applyingacupuncture treatments.

EfficacyThe end point adopted in this study was the

percentage of the patients with Z50% reduction inmigraine attack frequency. In the real acupuncture group,this parameter started to increase in the rst month oftreatment and the statistical signicant dierence betweengroups appeared in the second month of treatment(P=0.021). When the parameter adopted was thepercentage of the patients with Z40% reduction inmigraine attack frequency, the improvement in the realacupuncture group appeared in the rst month of thetreatment (P=0.044) and in the second month oftreatment (P=0.004). In the third month of thetreatment, the sham acupuncture group had a very highimprovement (Fig. 2). Consequently, at this point, thestatistically signicant dierence between groups disap-peared in all primary and secondary ecacy parametersadopted in this trial. In the rst 2 months of treatment thesham group improved only slightly (5.9% and 11.8%,respectively). Thus, if the sham group remained at thislevel until the end of the trial it could probably indicateno prophylactic activity or analgesic eects.44 The veryhigh improvement occurred in the sham acupuncturegroup in the third month of the treatment remainsunexplained to the authors. The most plausible hypoth-esis is that the result was due to chance. But, it would beimportant to observe in future trials if that kind of shamacupuncture called minimal acupuncture has anyretarded analgesic eect when applied in true acupuncturepoints as in this trial. The majority of acupuncture trialsin the sham group were done in false points.

The present trial showed that in the third month ofthe treatment 63.2% and 47.1% of the patients in the realand sham group, respectively, reached Z50% reductionin their migraine attacks frequency. Reduction inmigraine attacks reached was 48.1% and 44.1% in thereal and sham groups, respectively. The very highimprovement at the end of the treatment period in bothgroups is higher than it was shown in the majority of thetrials, which tested drugs for migraine prophylaxis. vander Kuy and Loman44 evaluated 22 relevant migraineprophylactic trials (double-blind and randomized) and noplacebo response was seen with more than 35% of thepatients who had Z50% reduction in the frequency ofmigraine attacks. When the reduction in the frequency ofmigraine attacks was analyzed, improvement in the

placebo setting was not observed above 40%. Therefore,eects noted in the sham group in this present trial dierfrom the expected for inert placebo3,4,26,27,45 inrandomized and blinded migraine trials.44 That improve-ment conrms the impression that sham acupuncturepromotes analgesic eects3,46,47 by unknown mechanism,probably with physiologic and psychologic components.

In this study, to protect the patients blinding, in thesham group 3 points were adopted in the head. Thesepoints are in the same nervous segment that was undertreatment, the head. Did it inuence the results? Thereis no scientic evidence for that eect. Several othernonspecic eects observed in trials could explain thehigh improvement of the sham group, such as thepatients belief in positive results of the acupuncturetreatment, the frequent patients contact with theacupuncturist doctor (16 visits in 3mo) and the magicpower of the nonconventional therapies have on people.In fact, the placebo eect is expected in any therapeuticintervention, independent of whether we are testing anactive drug or not, an active proceeding or not. Therefore,the placebo eects could be present in the real acupunc-ture group as well. However, it seems to be greater inpeople with pain48 particularly with invasive proce-dures.49 In migraine studies to test acute drugs such asanalgesics or triptans, the authors veried that theplacebo power changes depending on the access type ofthe intervention. They have seen that subcutaneousplacebo has higher ecacy (34%) than oral application(26%).49 Therefore, whether the placebo power ofthe invasive methods adopted as a control proceeding inthe present trial (minimal acupuncture) have increased theeects in the sham acupuncture group is unknown. Infact, the use of sham acupuncture as a control group maybe a greater problem in studies involving pain than inmeasuring other sort of symptoms.4

Silberstein and Goadsby50 stated that in controlledclinical trials to evaluate drugs in migraine prophylaxis,the ecacy is often rst noted in 4 weeks and continues toincrease for 3 months. In the analysis within the realacupuncture group, improvement in several variableswith statistical signicant dierences could be detected inthe rst month of treatment and it increased up to thethird (last) month of the treatment. In the shamacupuncture group, the improvement with statisticallysignicant dierence appeared only from the third monthof the treatment to the late posttreatment follow-upperiod. Therefore, these data indicate that the eect in thereal acupuncture group started faster and it increasedwith continued treatment.

The eects of the acupuncture groups (real andsham) decreased when the treatment was interrupted.Nevertheless, part of these improvements was preservedslightly in both groups through the early up to the lateposttreatment follow-up period. The long-lasting eectsof acupuncture intervention were observed in most recentand well-designed trials.2629,5153 This trial conrmsthe results of previous trials,54,55 which showed that nosignicant changes happened in the headache severity



with all kinds of acupuncture interventions either realor sham.

The analysis within each group showed improve-ment of all parameters evaluated except for the headacheseverity and vomiting in the real acupuncture group andheadache severity, nausea, and vomiting in the shamgroup. The improvement of nauseas appeared withstatistical signicant dierence in the analysis within thereal acupuncture group from the rst month of treatmentto the late posttreatment follow-up period. The acupunc-ture eect to control nausea is known and it has beendemonstrated in several high-quality clinical trials.4

The trial indicated of 2 dierent pathways fordevising adequate sample sizes in future essays. Theevidences found here in the improvement for both groupssupports the need for blinding patients undergoinginvasive sham procedures supports the requirements upto a 50% reduction. Then it would be necessary to have 2groups of at least 22 each one (total=44) to fulll therequirements of a 5% a and 20% b errors. To evaluatedierences in the third month of treatment, the estimatedsample size would need to grow sharply to at least 2groups of 146. Estimates for the late posttreatmentfollow-up are unstable due the large sampling variability.

Limitations of the StudyEnrollment of patients was dicult and most were

excluded, 286/359 (79.7%), because of the strict inclusionand exclusion criteria. The patients might have heardabout the benets of acupuncture and might haveexpected a good treatment response. Therefore, theinuence in the results of the patients belief and highexpectation on acupuncture treatment could not beexcluded.

Several factors could have contributed to the resultsfound in this trial. Few statistical signicant dierencesbetween real and sham acupuncture groups could be aconsequence of the presence of the great intergroupsvariability, the small size of the sample, the sort of controlgroup adopted as sham acupuncture in this trial (minimalacupuncture in true acupuncture points), and the strongplacebo eects identied in all kinds of the headachetrials.56

In conclusion, statistical signicant dierencesobserved between groups in some pain parametersevaluated were transient. Analysis within each groupdemonstrated that the sham and real acupuncture hadeects in almost all pain parameters evaluated and theywere maintained for a long time after the last acupuncturesession. The improvement in both groups was superior tothe expected placebo prophylactic eects with drugs. Thesmall size of the sample and the absence of another typeof control group did not allow for major conclusions. Weexpect that this trial contribute to reach a more adequatemethodology to achieve the response in the future aboutthe role acupuncture plays to prevent attacks in migrainesuerers.

ACKNOWLEDGMENTSThe authors thank Drs Lo Sz Hsien, Carlos

Moriyama, Norton Moritz Carneiro, and Isabel Giraltfor their helpful suggestions about acupuncture treatment.They thank Vanessa Monteiro Bugni, Ana Luisa VidigalSoares de Andrade, Heloisa de Lima Gomes, ErnaniAzevedo, and Gina Goulart for their help in organizingthe study and data, and especially Leda Fernandes foropening ways in the University Hospital. They also thankthe Clinic Hospital Public Relations and Press Assessor,Antonio Alberto Ravagnani and Isabel Cristina Gardenal,respectively, for their great help with the media. Theprincipal investigator wrote her study protocol, which ispart of the PhD thesis (Jerusa Alecrim-Andrade).

REFERENCES1. Filshie J, Cummings M. Acupuntura medica ocidental. In: Ersnt E,

White A, eds. Acupuntura-Uma Avaliacao Cientca. Barueri:Editora Manole Ltda; 2001:3976.

2. Boivie J, Brattberg G. Are there long lasting eects on migraineheadache after one series of acupuncture treatments? Am J ChinMed. 1987;15:6975.

3. Manias P, Tagaris G, Klementine K. Acupuncture in headache:a critical review. Clin J Pain. 2000;16:334339.

4. NIH Consensus Conference. Acupuncture. JAMA. 1998;280:15181524.

5. Ersnt E, White A. Conclusao. In: Ersnt E, White A, eds.Acupuntura-Uma Avaliacao Cientca. Barueri: Editora ManoleLtda; 2001:197203.

6. Manheimer E, White A, Berman B, et al. Meta-analysis: acupunc-ture for low back pain. Ann Internal Med. 2005;142:651663.

7. Stewart WF, Lipton RB, Celentano DD, et al. Prevalence ofmigraine headache in the United States. Relation to age, income,race, and other sociodemographic factors. JAMA. 1992;267:6469.

8. Chen Q. A brief introduction to migraine in the mainland of China.Cephalalgia. 1998;18(suppl 22):6264.

9. Lipton RB, Stewart WF. Prevalence and impact of migraine. NeurolClin. 1997;15:113.

10. Stewart WF, Lipton RB, Liberman J. Variation in migraineprevalence by race. Neurology. 1996;16:231238.

11. Garca E, Ristol A. Acupuntura y Neurologa. Rev Neurol.1997;25:894898.

12. von Peter S, Ting W, Scrivani S, et al. Survey on the use ofcomplementary and alternative medicine among patients withheadache syndromes. Cephalalgia. 2002;22:395400.

13. Laitinen J. Acupuncture for migraine prophylaxis: a prospectiveclinical study with six months follow-up. Am J Chin Med. 1975;3:271274.

14. Lenhard BS. Acupuncture in the prophylactic treatment of migraineheadaches: pilot study. N Z Med J. 1983;96:663666.

15. Boivie J, Brattberg G. An evaluation of acupuncture treatment formigraine. Acta Neurol Scand. 1984;69(suppl 98):268269.

16. Loh L, Nathan PW, Schott GD, et al. Acupuncture versus medicaltreatment for migraine and muscle tension headaches. J Neurol,Neurosurg, Psychiatry. 1984;47:333337.

17. Vincent CA. A controlled trial of the treatment of migraine byacupuncture. Clin J Pain. 1989;5:305312.

18. Hesse J, Mogelvang B, Simonsen H. Acupuncture versus metoprololin migraine prophylaxis: a randomized trial of trigger pointinactivation. J Internal Med. 1994;235:451456.

19. Baischer W. Acupuncture in migraine: long-term outcome andpredicting factor. Headache. 1995;35:472474.

20. Pintov S, Lahat E, Alstein M, et al. Acupuncture and the opioidsystem: implications in management of migraine. Pediatr Neurol.1997;17:129133.

21. Allais G, Lorenzo C, Quirico PE, et al. Acupuncture in theprophylactic treatment of migraine without aura: a comparison withunarizine. Headache. 2002;42:855861.

Alecrim-Andrade et al Clin J Pain Volume 24, Number 2, February 2008

104 r 2008 Lippincott Williams & Wilkins

22. Vickers A, Rees RW, Zollman CE, et al. Acupuncture for chronicheadache in primary care: large, pragmatic, randomized trial. BMJ.2004;328:744750.

23. Richardson PH, Vincent CA. Acupuncture for the treatment ofpain: a review of evaluative research. Pain. 1986;24:1540.

24. Melchart D, Linde K, Fischer P, et al. Acupuncture for recurrentheadaches: a systematic review of randomized controlled trials.Cephalalgia. 1999;19:779786.

25. Melchart D, Linde K, Berman B, et al. Acupuncture for idiopaticheadache (Cochrane Review). The Cochrane Library. Oxford:Update Software; 2005.

26. Linde K, Streng A, Jurgens S, et al. Acupuncture for patientswith migrainea randomized controlled trial. JAMA. 2005;293:21182125.

27. Melchart D, Streng A, Hoppe A, et al. Acupuncture in patientswith tension-type headache: randomized controlled trial. BMJ.2005;331:376379.

28. Alecrim-Andrade J, Maciel-Junior JA, Cladellas XC, et al.Acupuncture in migraine prophylaxis: a randomized sham-controlled trial. Cephalalgia. 2006;26:520529.

29. Diener HC, Kronfeld K, Boewing G, et al. Ecacy of acupuncturefor the prophylaxis of migraine: a multicentre randomisedcontrolled clinical trial. Lancet Neurol. 2006;5:310316.

30. International Headache Society. International Headache SocietyMembers Handbook 2000. Guidelines for Controlled Trials of Drugsin Migraine. Oxford: Scandinavian University Press; 1999:111133.

31. WHO. Guidelines for Clinical Research on Acupuncture. WesternPacic Series No. 15. Manila: WHO Regional Publications;1995:162.

32. IHS. Headache Classication Committee of the InternationalHeadache Society. Classication and diagnostic criteria for head-ache disorders, cranial neuralgias and facial pain. Cephalalgia.1988;8(7 suppl):196.

33. Daniel WD. Appendix IIStatistical Tables. BiostatisticsAFoundation for Analysis in the Health Sciences. New York, NY:John Wiley & Sons Inc; 1995:652.

34. Maciocia G. The Foundations of Chinese Medicine. Edimburg:Churchill Livingstone; 1989.

35. Nghi NV, Nguyen CR. Medicina Tradicional ChinesaAcupuntura-moxibustion & Masajes, 1. Barcelona: IBB S.A.; 1985.

36. Chenggu Y, Yi J, Biying H. Tratamiento de Las EnfermedadesMentales por Acupuntura y Moxibustion. Madrid: MiraguanoEdiciones; 1991.

37. Ross J. The Organ Systems of Traditional Chinese Medicine. 2nd ed.Edinburgh: Churchill Livingstone; 1985.

38. Liangye D, Yijun G, Xiaopin J, et al. Chinese Acupuncture andMoxibustion. 1st ed. Beijing: Foreign Languages Press; 1987.

39. Maciocia G. Headaches. In: Maciocia G, ed. The Practice of ChineseMedicine. London: Churchill Livingstone; 1994:154.

40. White A, Hayhoe S, Hart A, et al. Adverse events followingacupuncture: prospective survey of 32,000 consultations withdoctors and physiotherapists. BMJ. 2001;323:485486.

41. Rasmussen BK. Epidemiology and socio-economic impact ofheadache. Cephalalgia. 1999;25(suppl):2023.

42. Celentano DD, Linet MS, Stewart WF. Gender dierence in theexperience of headache. Soc Sci Med. 1990;30:12891295.

43. Centoze V, Polito BM, Valerio A, et al. Migraine with and withoutaura in the same patient: expression of a single clinical entity?Cephalalgia. 1997;17:585587.

44. van der Kuy P-HM, Lohman JJHM. A quantication of the placeboresponse in migraine prophylaxis. Cephalalgia. 2002;22:265270.

45. Ezzo J, Berman B, Hadhazy VA, et al. Is acupuncture eective forthe treatment of chronic pain? A systematic review. Pain. 2000;86:217225.

46. Streitberger K, Kleinhenz J. Introducing a placebo needle intoacupuncture research. Lancet. 1998;352:364365.

47. Macdonald JRA. Acupuncture analgesia and therapy. In: Wall PD,Melzack R, eds. Textbook of Pain. London: Churchill Livingstone;1994:906919.

48. Eikermann A, Diener HC. Eect of active treatment is lower whenusing placebo control in clinical trials on acute therapy of migraine.Cephalalgia. 2003;23:344347.

49. Diener HC. Placebo. Proceedings of Tenth Congress of theInternational Headache Society: IHC; 2001 June 29July 2001;New York City, New York, USA. Cephalalgia. 2001;21:248.

50. Silberstein SD, Goadsby PJ. Migraine: preventive treatment.Cephalalgia. 2002;22:491512.

51. Tavola T, Gala C, Conte G, et al. Traditional Chinese acupuncturein tension-type headache: a controlled study. Pain. 1992;48:325329.

52. Alecrim-Andrade J, Maciel-Junior JA, Cladellas XC, et al. Thelong-lasting eects of acupuncture treatment for migraine prophy-laxis: 44 weeks post-treatment follow-up. Cephalalgia. 2005;25:942943.

53. Hsieh LLC, Kuo CH, Lee LH, et al. Treatment of low backby acupressure and physical therapy: randomized controlled trial.BMJ. 2006;332:696700.

54. White AR, Resch KL, Chan JCK, et al. Acupuncture for episodictension-type headache: a multicentre randomized controlled trial.Cephalalgia. 2000;20:632637.

55. Karst M, Reinhard M, Thum P, et al. Needle acupuncturein tension-type headache: a randomized, placebo-controlled study.Cephalalgia. 2001;21:637642.

56. Diener HC. Placebo in headache trials. Editorial. Cephalalgia.2003;23:485.



Documents

Acupuncture in Migraine Prevention- A Randomized Sham Controlled Study With 6-Months Posttreatment Follow-up